Optimizing Computations and Allocations on High Performance and Cloud Computing Systems

Graduate Thesis Or Dissertation

Citations

Citeable URL: https://scholar.colorado.edu/concern/graduate_thesis_or_dissertations/g732d925b

Abstract

Over the last decade, many research and development projects have focused on Cloud Computing systems. After forming around the early research papers and the first commercial cloud offerings in 2006-2008, the field has seen a tremendous progress and has provided the primary infrastructure and technology for applications at small, medium, and large scales. Cloud Computing systems have provided diverse on-demand resources to individual researchers and developers, groups and entire institutions, as well as commercial companies and government organizations. Clouds have also found their niche in scientific computing applications, offering attractive alternatives to High Performance Computing models and systems. While cloud economics and technologies have significantly matured recently, there is much active research revolving around topics such as optimality, usability, manageability, and reproducibility in the latest studies. This dissertation presents our findings and relevant developments at the intersection of Cloud Computing and such “flavors” of computing as High Performance Computing and High Throughput Computing. We primarily focus on optimality issues in this area and propose solutions that address the needs of individual researchers and research groups with limited computational and financial resources.

Creator